Comments (4)
hungbie
commented on July 29, 2024
What you are refering to is a Binary Classification problem and hence using sigmoid with binary cross entropy is possible. Thus if you are using that you only need to have 1 class in the final prediction.
However here it is indeed still a binary problem but they are using categorical cross entropy and is compatible with more than 2 classes. And since categorical cross entropy is used they need to specify that there are 2 classes, hence 2.
from retina-unet.
Kevin-Moon
commented on July 29, 2024
Thank you for your kind response.
Then, do you think that they have almost the same result in terms of
accuracy or performance?
Thanks in advance.
Best
Kevin
2018. 2. 12. 오후 6:22에 "Pham Tan Hung" <[email protected]>님이 작성:
… What you are refering to is a Binary Classification problem and hence
using sigmoid with binary cross entropy is possible. Thus if you are using
that you only need to have 1 class in the final prediction.
However here it is indeed still a binary problem but they are using
categorical cross entropy and is compatible with more than 2 classes. And
since categorical cross entropy is used they need to specify that there are
2 classes, hence 2.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#45 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AYAAVo61Eue0VTWP0PQNA4DUThVO5OPEks5tUAK3gaJpZM4RwE_v>
.
from retina-unet.
hungbie
commented on July 29, 2024
For this problem yes they should be the same. However say we now want to also detect the ONH and have 3 classes instead then we can no longer use Sigmoid with Binary Cross Entropy.
from retina-unet.
Kevin-Moon
commented on July 29, 2024
I got it!!
Thank you so much.
I hope that you will achieve great work.
Best
Kevin
2018. 2. 12. 오후 7:13에 "Pham Tan Hung" <[email protected]>님이 작성:
… For this problem yes they should be the same. However say we now want to
also detect the ONH and have 3 classes instead then we can no longer use
Sigmoid with Binary Cross Entropy.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#45 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AYAAVugIhZWeoyBFlCx4kJspwo4O1pDtks5tUA7PgaJpZM4RwE_v>
.
from retina-unet.
Related Issues (20)
- Produce the segmentation for a whole image.
- only the configuration.txt file is available in the resulting test folder
- Why is there no 1st_manual and 2nd_manual under test file after I download the DIRVE? HOT 4
- Could not get the files after running training code HOT 3
- can't find sample_input_imgs
- how to train on other database?
- how to train on my own database HOT 1
- Keras and tf version HOT 2
- TypeError: float() argument must be a string or a number, not 'TiffImageFile' HOT 2
- ImportError: cannot import name 'jaccard_similarity_score' HOT 1
- how to create overlapping patches from images
- version
- How to python run_training.py ?
- please help "ImportError: No module named visualize_util
- For those who are interested in the theoretical part of this code... the article's title of this code is "Retina Blood Vessel Segmentation Using A U-Net Based Convolutional Neural Network".
- Why i am getting less performance than yours?
- Process finished with exit code -1073740791 (0xC0000409) HOT 1
- Trouble with model architecture HOT 1
- Any subsequent image repair work
- semi-supervised learning
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from retina-unet.